TlEI\[ArvKABLE SPECTES OF ECIirurtOIDS. 47 



consisting of cubicnl or cylindrical cells arranged in a row. 

 What then appear like cells are in reality cross-sections of 

 sarcolemma-like sheaths that inclose each a bundle of longitudi- 

 nal muscle fibers {Ijii.). In most places this layer forms ap- 

 parently the innermost wall of the median vessel. A continuous 

 endothelium could not be brought into view. However, there 

 exist, here and there on the internal surface of the wall, peculiar 

 isolated cells of irregularly stellate shape (ent.c.) They are fixed 

 in position in that a part of their body is inserted in the inter- 

 stice between the sarcolemma-like sheaths. If the cells can 

 really be considered to constitute a part of the wall, they may 

 possibly represent endothelial cells. In the proboscis of Bonellia 

 7)iisakiensis, which I studied fer the sake of comj)arison, I found 

 the median vessel as well as the lateral vessels provided with 

 distinct endothelium. According to Kietsch, neither an endothel- 

 ium nor a layer of longitudinal muscle bundles, such as I have 

 described, should exist in the wall of the median vessel in the 

 proboscis of Thalassemia neptuni, Echiurus unicinctus and Bonel- 

 lia minor. 



3. TJialassenia elef/ans I. Ikeda. 



(Fig. 4, PI. I. ; figs. 48 and 49, PI. IV.) 



1904. Th. elerjans, I. Ikeda, Jour. Sei. Coll., Vol. XX., Art. 4, p. 65. 



During July, 1902, I have obtained in all four specimens 

 of this highly interesting species. At first it was thought they 

 nn'ght possibly be the male of Th. tœnioides, but a closer ex- 

 amination soon revealed the fact that I had before me a new 



